2022 Fiscal Year Final Research Report
Development of a Self Machine Learning System for Fast Detection of Intrusions to Next Generation in-Vehicle Network
| Project/Area Number |
19K11881
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 60040:Computer system-related
|
|---|
| Research Institution | Hiroshima City University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2019-04-01 – 2023-03-31
|
| Keywords | プログラマブルシステム / ネットワークセキュリティ / ネットワーク侵入検知システム / 機械学習 / 決定グラフ / 論理設計 / 多値論理 |
|---|
| Outline of Final Research Achievements |
In this study, we developed an in-vehicle network intrusion detection system (NIDS) and its computer aided design tools. The developed system is based on a hybrid type using signature-based detection and anomaly detection in a complementary manner. In the system, both signature-based detection and anomaly detection are realized with tree-based hardware, and thus, it achieves high-speed detection and compact size. Since we have already developed a fast circuit for signature-based detection that can achieve 11 to 23 times faster than software-based NIDSs in our previous study, we focused on development of a circuit for anomaly detection in this study. By making tree-based hardware multiple-valued, we confirmed that it is possible to shorten processing time in 40% and reduce size in 20%.
|
| Free Research Field |
情報学
|
| Academic Significance and Societal Importance of the Research Achievements |
本研究では，機械学習（いわゆるAI）によって，車載ネットワークへのハッキングなどの不正侵入を高速に検知するシステムを開発した．AIでの計算部分に多値決定グラフの技術を用いることで，処理の高速化だけでなく，車載などの組込み機器に必須となるサイズ削減も達成できた．新たな侵入手口にも柔軟に対応できる仕組みも備えているため，機器へのネットワークを介した不正侵入における安全性と利便性の両立が可能になった．当研究では，システムの設計・最適化手法も合わせて開発したため，車載機器に限らず，様々なIoT機器に応用することができ，当該技術の応用が進めば，より安全で快適なIoT社会が実現されるだろう．
|
